How could I calculate the value counts within a string col?

df col
0 fruit["apple"], colour["green", "yellow" ] 
1 colour["yellow"] 
2 colour["brown"] 

Expected Output

fruit  1
colour 3

CodePudding user response：

Use Series.str.extractall with substrings joined by | for regex or:

s = df['col'].str.extractall('(fruit|colour)')[0].value_counts()
print (s)
colour    3
fruit     1
Name: 0, dtype: int64

Or get words before [ for more dynamic solution:

s = df['col'].str.extractall(r'(\w )\[')[0].value_counts()
print (s)
colour    3
fruit     1
dtype: int64